My fascination with the brain and my insatiable curiosity about all it does and is responsible for pushed me towards neuroscience research.
For the last 9 years, I have been a neuroscience researcher applying cutting-edge research techniques in behaving rodents to specifically target, manipulate, and record from specific neuron populations in specific regions of the brain.
I was so excited to be able to ask questions about how the brain works and to then design experiments to test them. To gain some meaningful insight that I can then share with the scientific community so that we can better understand our own behavior as humans, and to provide new insights that can be used in creating new therapeutic treatments for real-world problems, like alcoholism.
When we had a new hypothesis that we wanted to test about the roles of stress neurons of the amygdala in the escalation of binge-drinking that leads to alcoholism (that we were not yet set up for), we had to do the following course of events:
1) [~ 1 Week] Write a new experimental protocol to submit for approval. This involved laying out the experiment’s aims and hypothesis, what treatments and drugs and behavior the mice would undergo. But this also involves justifying experiments can only be performed in rodents, calculating the # of mice required for the new experiment, the number of breeding cages and offspring and mating schemes to produce those mice, flow charts, detailed descriptions of all people involved and their experience levels.
2) [~2-6 Weeks] This protocol is then submitted to the Institutional Animal Care and Use Committee (IACUC), which only meets once per month, and which must approve all experiments before we can even begin setting up our mouse colonies. Chances are that the IACUC committee will disagree with some aspect of your protocol and require an edited version to their specifications. After they finally meet AGAIN, we are granted approval and can start our experiments…
3) [12 weeks] …except that our experiments require a specific strain of mice, which we then need to purchase breeding pairs from a national animal repository, which would arrive in a week and then undergo a required 4 week-long quarantine before they can be accessed. We then set up male and female breeding pairs and wait for them to mate, produce litters of pups, and wait for them to reach an appropriate age and size (which take ~2 months to produce the first mice we can use.)
4) [4 weeks] All of the research techniques we employed used special engineered proteins packed into otherwise-inert viral vectors, which allowed us to express our special proteins in specific genetically-targeted types of neurons within specific regions of the brain. These proteins would let us either excited or inhibit our targeted neurons using laser light delivered via an optical fire directly into the brain.Generally, we had to wait 4 weeks after surgery before our engineered proteins had been produced in high enough quantities for us to begin our experimental-manipulations.
5) [3-5 weeks] Our experiments then aimed to study how specific neurons of the amygdala altered binge-drinking of alcohol in mice (corticotrophin-factor releasing neurons of the central amygdala, to be exact). Mice do not readily drink alcohol and it would take about 3-5 weeks of daily 4-hour-long sessions of alcohol access before the mice reached a stable baseline and we could start manipulating the targeted neurons to test our hypothesis. There were only 8-behavior boxes available in our lab, meaning we could only have 8 mice undergoing the experiment at a time (4 control mice and 4 experimental).
6) [5 weeks] We then spent 1 week doing our exerimental-manipulations using laser-light excitation during drinking behavior, collecting data on changes in licking patterns, another week of drinking without manipulations, then 3 weeks of abstinence form alcohol, then two more weeks of alcohol drinking.
7) [1 week] Once the experiments have been performed and the data collected, we have to verify that our virus actually expressed the protein it was supposed to in the right location, and that our implants were actually shining light at that location. The brains had to be extracted from the mice, preserved in fixative, frozen, sliced, mounted and then visualized on a confocal microscope, one slice at a time, to see if we can see the fluorescent yellow of our protein, and if it seems to be in the correct location.
[!] Uh-oh… one of our mice didn’t have strong viral expression and one of them had bad implant placement… thats 2 out of our 8 mice… and all of their data must be eliminated Wait, what?? Oh, no… they were both the experimental group! So now we only have data from 2 experimental mice and 4 control mice to test our hypothesis, which is statistically impossible.
We first thought of our new hypothesis that we desired to test 32 weeks (8 months) ago… and we still can’t conclude anything.
As much as the brain fascinated me and how much I wanted to understand it, my enthusiasm for research was drained out of me. The idea of hypothesis testing became lost in the sea of logistical execution. You can start to see how scientists may be subconsciously biased towards getting a result. How could we not, when we invest so much of ourselves for so long into one question? By the end, I was afraid to ask a new question….or to question the statistical power of my data…
So when you ask me, why study data science? My answer is simple:
I can form a hypothesis about relationships between real-world variables, find and clean massive amounts of data from multiple databases, perform deep exploratory data analyses, perform a myriad of statistics and data modeling, come to a conclusion and prepare a report with supporting figures and within what… a few days? a week? 4 weeks? That’s 32 weeks, vs 1-4 weeks.
I can think of a brand-new question or relationship that I’ve never considered before and have no prior data about and reach a meaningful conclusion that can help guide real-world people and decision-making immediately…
…whereas my neuroscience research may contribute to some meaningful insights about the amygdala’s role in binge drinking, when combined with accumulating research over the next several years, which* may* provide information for a scientist to think of a new drug that targets the neural circuitry that may have the desired effect, that needs to be tested in rodents and then in primates. If it works and if a pharmaceutical company buys the license to produce the drug and to conduct several years’ worth of clinical trials (assuming it survives the process) before it might help real-world people…
Yeah…I’ll take the future career in data science, please…
( No offense to my former colleagues. )